Using Twitter Sentiment and Q-Learning to Trade Stocks

Summary

This study examines whether daily Twitter sentiment can help predict stock returns and support trading decisions. It combines machine learning to assess the sentiment signal with reinforcement learning, specifically Q-learning, to derive a trading policy from that signal.

The reported results suggest sentiment is more predictive for companies whose prices reflect expectations of future growth, and around major events that attract public attention. The Q-learning strategy is reported to outperform a strategy based on machine-learning predictions. The document provides no sample details, performance figures, transaction-cost analysis, or discussion of how the strategies were validated, so the strength and real-world applicability of the results cannot be assessed from this description alone.

Key ideas

• Daily Twitter sentiment is evaluated as a signal for predicting stock returns.
• Q-learning is used to turn the sentiment signal into a trading policy.
• The reported predictive value is stronger for growth-expectation-driven stocks and around major public events.
• The reinforcement-learning strategy reportedly outperforms trading directly from machine-learning predictions.

Tags

# Trading the Twitter Sentiment with Reinforcement Learning


# Trading the Twitter Sentiment with Reinforcement Learning









This paper is to explore the possibility to use alternative data and artificial intelligence techniques to trade stocks. The efficacy of the daily Twitter sentiment on predicting the stock return is examined using machine learning methods. Reinforcement learning(Q-learning) is applied to generate the optimal trading policy based on the sentiment signal. The predicting power of the sentiment signal is more significant if the stock price is driven by the expectation of the company growth and when the company has a major event that draws the public attention. The optimal trading strategy based on reinforcement learning outperforms the trading strategy based on the machine learning prediction.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.